*3.1. Optimisation and Validation of the CRISPR-Cas9 System in L. braziliensis*

To test the feasibility and efficiency of sgRNA-guided, Cas9-mediated gene editing in *L. braziliensis*, we first targeted an integrated transgene coding for green fluorescent protein (eGFP). To this end, we generated a stable cell line of *L. braziliensis* expressing Cas9 and T7 RNAP from an episome (pTB007). The eGFP coding sequence was fused into the pIR-mcs3+ plasmid [53], and the linearised plasmid was transfected into *L. braziliensis*, leading to integration into the small subunit rRNA (18S) coding sequence (Figure 1A).

We confirmed the expression of Cas9 protein by Western blot analysis (Figure S1A) and the detection of *T7RNAP* mRNA by qRT-PCR (Figure S1B). To better assess the efficiency of CRISPR–Cas9-mediated gene editing in *L. braziliensis*, we included Old World *L. donovani* strain 1S for comparative purposes, since the latter has long been used as a model for homologous recombination and genetic complementation in our laboratory.

**Figure 1.** CRISPR–Cas9-mediated disruption of *eGFP* gene as proof-of-principle test in *L. braziliensis*. (**A**) Generation of Cas9–eGFP-expressing parasites. *Left panel*: plasmid pTB007 [17] bearing *hSpCas9* and *T7 RNAP* transgenes was transfected as circular episome into *L. braziliensis* PER005cl2 wild-type parasites. Transfectants were selected under Hygromycin B pressure. *Right panel*: schematic depiction of the double cross-over homologous recombination strategy to integrate the linearised pIR–*eGFP* construct into the SSU rRNA locus of *L. braziliensis* Cas9-expressing parasites. Regions shown are the SSU rRNA sequences on either ends resulting from *Swa*I restriction digest, the *eGFP* ORF, and the nourseothricine resistance gene ORF (*SAT*, encoding streptothricin-acetyltransferase). (**B**) Schematic representation of the *eGFP* locus and locations of the six 20-nt guide RNA sequences used for gene disruption; the guide sequence pairs with the DNA target (blue bar), directly upstream of a requisite 5'-NGG-3' adjacent motif (PAM). The green arrowhead indicates the predicted Cas9 cleavage sites. Only the coding strand is shown. Binding sites of primers used for genotyping of genetically engineered parasites are denoted by arrows. The PCR fragment size depended on the pair of single guide RNAs (sgRNAs) tested. Sets of sgRNAs tested: set 1 = *eGFP*-52-5'sgRNA and *eGFP*-253-3'sgRNA; set 2 = *eGFP*-52-5'sgRNA and *eGFP*-612-3'sgRNA; set 3 = *eGFP*-52-5'sgRNA and *eGFP*-639-3'sgRNA; set 4 = *eGFP*-553-5'sgRNA and *eGFP*-639-3'sgRNA; set 5 = *eGFP*-378-5'sgRNA and *eGFP*-612-3'sgRNA; set 6 = *eGFP*-378-5'sgRNA and *eGFP*-639-3'sgRNA. (**C**) Flow cytometry analysis of eGFP–Cas9-expressing parasites before and after transfection of *eGFP*-targeting sgRNAs. Efficiency of *eGFP* disruption using 6 different sets of sgRNAs in *L. donovani* (left panel) and *L. braziliensis* (right panel) as quantified by GFP expression. Each set of two sgRNAs was co-transfected with two donor DNAs; transfections were done in triplicate. Sets of sgRNAs tested (labelled as set 1 to 6 in the graphs) consisted of pairs as described in Figure 1B. P, parental cell line Cas9/T7/eGFP. The gating scheme, a representative histogram, and all FACS plots showing the percentage of GFP-positive cells are shown in Supplemental Figures S2 and S3. 175

The *L. braziliensis* and *L. donovani* parental cell lines (Cas9/T7/GFP) were co-transfected with a pair of *eGFP*-targeted sgRNAs and corresponding donor DNA cassettes (i.e., homologous repair templates) to facilitate homology-directed repair [54,55]. Six different sets of dual sgRNAs and donor DNAs (Figure 1B; Table S1) were tested in triplicate. Transfectants were subjected to blasticidin and puromycin drug selection. At this point, drug selection (hygromycin B) for maintenance of the pTB007 episome encoding Cas9 and T7 RNAP and nourseothricine selection for the integrated pIR-mcs-*eGFP* were stopped.

In *L. donovani* 1S, the antibiotic selection pressure with the drug-selectable markers was kept constant throughout the selection period (10 µg/mL blasticidin, 25 µg/mL puromycin), following the optimised conditions established previously for this parasite strain in our group (data not shown). Survival of *L. donovani* double drug-resistant transfectants became apparent 6–10 days after transfection. Transfectants with *eGFP*-targeted sgRNAs set 5 and set 6 were the first to emerge in culture (6 and 9 days after transfection, respectively). Candidate *eGFP* replacement populations were passaged at least twice before analysing the gene disruption outcome by flow cytometry. Each of the 6 pairs of sgRNAs resulted in highly efficient reduction of GFP expression (Figure 1C, left panel; Figure S2). PCR analysis of genomic DNA with primers amplifying the entire *eGFP* ORF showed no detectable band corresponding to the *eGFP* transgene in all selected *L. donovani* lines, but bands of higher size appeared, indicating the integration of the donor repair cassettes (Figure S4B, left panel), as expected (Figure S4A). This was verified with *BSD* and *PAC* gene-specific primers (Figure S4B, left panel) and confirmed the high efficiency of CRISPR–Cas9-mediated *eGFP* disruption in *L. donovani*.

In *L. braziliensis* PER005cl2 we first established the suitable concentrations of antibiotic selection through titration curves for 7 days (Figure S5). On this basis we decided to subject the parasites at first to the lowest concentrations of antibiotics that had a growth inhibitory effect, i.e., blasticidin at 2.5 µg/mL (~IC85) and puromycin at 10 µg/mL (~IC65). The first *L. braziliensis* drug-resistant transfectants to emerge in culture, as in *L. donovani*, were those transfected with *eGFP*-targeted sgRNAs set 5 (12–14 days after transfection) and set 6 (14 days after transfection). Transfectants with the other *eGFP* sgRNA sets (1, 2, 3 and 4) emerged 18–22 days after transfection. Candidate *eGFP* replacement populations were passaged at least twice and then analysed by flow cytometry as non-clonal populations. By flow cytometric analysis, sgRNAs sets 5 and 6 were the most efficient to abrogate the eGFP expression (0.02–4.30% GFP-positive cells), whereas sgRNA set 3 was slightly less efficient (0.69–11.4% GFP-positive cells). The sgRNAs sets 1, 2 and 4 were the least efficient (2.91–47.00% GFP-positive cells) (Figure 1C, right panel; Figure S3). Genomic DNAs from these parasite populations were examined by PCR confirming a complete loss of the *eGFP* transgene only in three selected *L. braziliensis* lines (*eGFP*-null mutants 5.1, 5.3 and 6.3) (not shown), which were transfected with the most potent sgRNAs, sets 5 and 6. For the other selected *L. braziliensis* lines, a band corresponding to the unmodified *eGFP* gene was still detected with varying intensities (Figure S4, right panel, for *eGFP* mutants 3.1, 3.2, and 3.3). PCR analysis with *eGFP* gene-specific primers also showed bands of higher size indicating the integration of the donor repair cassettes in the *L. braziliensis eGFP* mutants (Figure S4B, right panel), as expected (Figure S4A). While the blasticidin replacement cassette was confirmed to be integrated in all *L. braziliensis* selected lines by PCR analysis with *BSD*-specific primers (Figure S4B, right panel), the puromycin replacement cassette was detected in twelve out of 18 selected *L. braziliensis* lines, as assessed using *PAC*-specific primers (Figure S4B, right panel). This outcome reflected the moderate antibiotic selective pressure used to generate the *L. braziliensis eGFP* mutants.

At day 35 after transfection of the *L. braziliensis* Cas9/T7/eGFP parental cell line, inspection of the two *L. braziliensis* mock-transfected controls showed minimal growth. To impose a more stringent dual antibiotic selection, the mock cultures and selected *eGFP* mutants were passaged in complete M199 medium with blasticidin at 5 µg/mL (~IC99.7) and puromycin at 20 µg/mL (~IC96). The mock-transfected cultures succumbed to the antibiotic pressure within 4 days, while the *eGFP* mutant populations proliferated. This double antibiotic selection regimen was used in all subsequent experiments.
